Context-free and context-sensitive dynamics in recurrent neural networks

Continuous-valued recurrent neural networks can learn mechanisms for processing context-free languages. The dynamics of such networks is usually based on damped oscillation around fixed points in state space and requires that the dynamical components are arranged in certain ways. It is shown that qualitatively similar dynamics with similar constraints hold for a(n)b(n)c(n), a context-sensitive language. The additional difficulty with a(n)b(n)c(n), compared with the context-free language a(n)b(n), consists of 'counting up' and 'counting down' letters simultaneously. The network solution is to oscillate in two principal dimensions, one for counting up and one for counting down. This study focuses on the dynamics employed by the sequential cascaded network, in contrast to the simple recurrent network, and the use of backpropagation through time. Found solutions generalize well beyond training data, however, learning is not reliable. The contribution of this study lies in demonstrating how the dynamics in recurrent neural networks that process context-free languages can also be employed in processing some context-sensitive languages (traditionally thought of as requiring additional computation resources). This continuity of mechanism between language classes contributes to our understanding of neural networks in modelling language learning and processing.

Simultaneous steady state and dynamic design of process systems using structural indicators

The simultaneous design of the steady-state and dynamic performance of a process has the ability to satisfy much more demanding dynamic performance criteria than the design of dynamics only by the connection of a control system. A method for designing process dynamics based on the use of a linearised systems' eigenvalues has been developed. The eigenvalues are associated with system states using the unit perturbation spectral resolution (UPSR), characterising the dynamics of each state. The design method uses a homotopy approach to determine a final design which satisfies both steady-state and dynamic performance criteria. A highly interacting single stage forced circulation evaporator system, including control loops, was designed by this method with the goal of reducing the time taken for the liquid composition to reach steady-state. Initially the system was successfully redesigned to speed up the eigenvalue associated with the liquid composition state, but this did not result in an improved startup performance. Further analysis showed that the integral action of the composition controller was the source of the limiting eigenvalue. Design changes made to speed up this eigenvalue did result in an improved startup performance. The proposed approach provides a structured way to address the design-control interface, giving significant insight into the dynamic behaviour of the system such that a systematic design or redesign of an existing system can be undertaken with confidence.

Minimum-order stable recursive filter design via the genetic algorithm approach

In this paper, the minimum-order stable recursive filter design problem is proposed and investigated. This problem is playing an important role in pipeline implementation sin signal processing. Here, the existence of a high-order stable recursive filter is proved theoretically, in which the upper bound for the highest order of stable filters is given. Then the minimum-order stable linear predictor is obtained via solving an optimization problem. In this paper, the popular genetic algorithm approach is adopted since it is a heuristic probabilistic optimization technique and has been widely used in engineering designs. Finally, an illustrative example is sued to show the effectiveness of the proposed algorithm.

The relationship between radiation-induced chemical processes and transverse relaxation times in polymer gel dosimeters

The effects of ionizing radiation in different compositions of polymer gel dosimeters are investigated using FT-Raman spectroscopy and NMR T-2 relaxation times. The dosimeters are manufactured from different concentrations of comonomers (acrylamide and N,N'-methylene-bis-acrylamide) dispersed in different concentrations of an aqueous gelatin matrix. Results are analysed using a model of fast exchange of magnetization between three proton pools. The fraction of protons in each pool is determined using the known chemical composition of the dosimeter and FT-Raman spectroscopy. Based on these results, the physical and chemical processes in interplay in the dosimeters are examined in view of their effect on the changes in T-2 The precipitation of growing macroradicals and the scavenging of free radicals by gelatin are used to explain the rate of polymerization. The model describes the changes in T-2 as a function of the absorbed dose up to 50 Gy for the different compositions. This is expected to aid the theoretical design of new, more efficient dosimeters, since it was demonstrated that the optimum dosimeter (i.e, with the lowest dose resolution) must have a range of relaxation times which match the range of T-2 values which can be determined with the lowest uncertainty using an MRI scanner.

The advantages of theft over toil: The design inference and arguing from ignorance

Intelligent design theorist William Dembski has proposed an explanatory filter for distinguishing between events due to chance, lawful regularity or design. We show that if Dembski's filter were adopted as a scientific heuristic, some classical developments in science would not be rational, and that Dembski's assertion that the filter reliably identifies rarefied design requires ignoring the state of background knowledge. If background information changes even slightly, the filter's conclusion will vary wildly. Dembski fails to overcome Hume's objections to arguments from design.

A time-harmonic inverse methodology for the design of RF coils in MRI

An inverse methodology is described to assist in the design of radio-frequency (RF) coils for magnetic resonance imaging (MRI) applications. The time-harmonic electromagnetic Green's functions are used to calculate current on the coil and shield cylinders that will generate a specified internal magnetic field. Stream function techniques and the method of moments are then used to implement this theoretical current density into an RF coil. A novel asymmetric coil operating for a 4.5 T MRI machine was designed and constructed using this methodology and the results are presented.

A method for the design of MRI radiofrequency coils based on triangular and pulse basis functions

This paper presents a numerical technique for the design of an RF coil for asymmetric magnetic resonance imaging (MRI) systems. The formulation is based on an inverse approach where the cylindrical surface currents are expressed in terms of a combination of sub-domain basis functions: triangular and pulse functions. With the homogeneous transverse magnetic field specified in a spherical region, a functional method is applied to obtain the unknown current coefficients. The current distribution is then transformed to a conductor pattern by use of a stream function technique. Preliminary MR images acquired using a prototype RF coil are presented and validate the design method. (C) 2002 Elsevier Science B.V. All rights reserved.

The VITATOPS (vitamins to prevent stroke) Trial: Rationale and design of an international, large, simple, randomised trial of homocysteine-lowering multivitamin therapy in patients with recent transient ischaemic attack or stroke

Background: Epidemiological studies suggest that raised plasma concentrations of total homocysteine (tHcy) may be a common, causal and treatable risk factor for atherothromboembolic ischaemic stroke. Although tHcy can be lowered effectively with small doses of folic acid, vitamin B-12 and vitamin B-6, it is not known whether lowering tHcy, by means of multivitamin therapy, can prevent stroke and other major atherothromboembolic vascular events. Purpose: To determine whether vitamin supplements (folic acid 2 mg, B-6 25 Mg, B-12 500 mug) reduce the risk of stroke, and other serious vascular events, in patients with recent stroke or transient ischaemic attacks of the brain or eye (TIA). Methods: An international, multi-centre, randomised, double-blind, placebo-controlled clinical trial. Results: As of November 2001, more than 1,400 patients have been randomised from 10 countries in four continents. Conclusion: VITATOPS aims to recruit and follow up 8,000 patients between 2000 and 2004, and provide a reliable estimate of the safety and effectiveness of dietary supplementation with folic acid, vitamin B-12, and vitamin B-6 in reducing recurrent serious vascular events among a wide range of patients with TIA and stroke. Copyright (C) 2002 S. Karger AG, Basel.

NMR shim coil design utilising a rapid spherical harmonic calculation method

A rapid spherical harmonic calculation method is used for the design of Nuclear Magnetic Resonance shim coils. The aim is to design each shim such that it generates a field described purely by a single spherical harmonic. By applying simulated annealing techniques, coil arrangements are produced through the optimal positioning of current-carrying circular arc conductors of rectangular cross-section. This involves minimizing the undesirable harmonies in relation to a target harmonic. The design method is flexible enough to be applied for the production of coil arrangements that generate fields consisting significantly of either zonal or tesseral harmonics. Results are presented for several coil designs which generate tesseral harmonics of degree one.

Evaluating design proposals for complex systems with work domain analysis

In this paper we propose a new framework for evaluating designs based on work domain analysis, the first phase of cognitive work analysis. We develop a rationale for a new approach to evaluation by describing the unique characteristics of complex systems and by showing that systems engineering techniques only partially accommodate these characteristics. We then present work domain analysis as a complementary framework for evaluation. We explain this technique by example by showing how the Australian Defence Force used work domain analysis to evaluate design proposals for a new system called Airborne Early Warning and Control. This case study also demonstrates that work domain analysis is a useful and feasible approach that complements standard techniques for evaluation and that promotes a central role for human factors professionals early in the system design and development process. Actual or potential applications of this research include the evaluation of designs for complex systems.

In situ measures of spawning synchrony and fertilization success in an intertidal, free-spawning invertebrate

The rocky intertidal zone has the potential to be one of the harshest environments for free-spawning organisms, but empirical data on fertilization success are scarce. Here, I report on an intertidal, solitary ascidian, Pyura stolonifera, which was observed to spawn at low tide. At a scale likely to be most important to gametes (metres, duration of tide), approximately 30% of individuals in the population were spawning synchronously. Spawned gametes remained in a viscous matrix and this appeared to minimise their dilution. Fertilization success varied greatly among individuals (0 to 92%) and was related to the distance to the nearest neighbouring spawner. Occasional wave wash facilitated the movement of sperm between spawners. Fertilization success in some individuals was limited by the scarcity of sperm whilst the experimental addition of sperm did not increase success in others.

Hair morphine concentrations of fatal heroin overdose cases and living heroin users

Aims To compare heroin and other opiate use of heroin overdose fatalities, current street heroin users and drug-free therapeutic community clients. Design Hair morphine concentrations that assess heroin use and other opiate use in the 2 months preceding interview or death were compared between heroin overdose fatalities diagnosed by forensic pathologists (fOD) (n = 42), current street heroin users (CU) (n = 100) and presumably abstinent heroin users in a drug-free therapeutic community (TC) (n = 50). Setting Sydney, Australia. Findings The mean age and gender breakdown of the three samples were 32.3 years, 83% male (FOD), 28.7 years, 58% male (CU) and 28.6 years, 60% male (TC). The median blood morphine concentration among the FOD cases was 0.35 mg/l, and 82% also had other drugs detected. There were large differences between the three groups in hair morphine concentrations, with the CU group (2.10 ng/mg) having concentration approximately four times that of the FOD group (0.53 ng/mg), which in turn had a concentration approximately six times that of the TC group (0.09 ng/mg). There were no significant differences between males and females in hair concentrations within any of the groups. Hair morphine concentrations were correlated significantly with blood morphine concentrations among FOD cases (r = 0.54), and self-reported heroin use among living participants (r = 0.57). Conclusions The results indicate that fatal cases had a lower degree of chronic opiate intake than the active street users, but they were not abstinent during this period.

Ecological interface design for pasteurizer II: A process description of semantic mapping

Ecological interface design (EID) is proving to be a promising approach to the design of interfaces for complex dynamic systems. Although the principles of EID and examples of its effective use are widely available, few readily available examples exist of how the individual displays that constitute an ecological interface are developed. This paper presents the semantic mapping process within EID in the context of prior theoretical work in this area. The semantic mapping process that was used in developing an ecological interface for the Pasteurizer II microworld is outlined, and the results of an evaluation of the ecological interface against a more conventional interface are briefly presented. Subjective reports indicate features of the ecological interface that made it particularly valuable for participants. Finally, we outline the steps of an analytic process for using EID. The findings presented here can be applied in the design of ecological interfaces or of configural displays for dynamic processes.

Superhyperfine interactions in inhomogeneously broadened paramagnetic centers observed via a hole-burned free induction decay

Superhyperfine interactions in inhomogeneously broadened paramagnetic centers are observed using a single high-turn-angle microwave pulse. The free induction signal that follows the hole-burning pulse exhibits oscillations that are distinct from the oscillatory free induction decay observable in some inhomogeneously broadened systems. It contains frequencies characteristic of the superhyperfine splittings, together with a zero frequency component. Experimental examples of the effect in both orientationally disordered (powdered) and structurally disordered (glassy) systems are presented and compared with the conceptually similar Fourier transform electron paramagnetic resonance detected nuclear magnetic resonance experiment, together with numerical simulations. (C) 2003 American Institute of Physics.

Genetic algorithms for design of liquid retaining structure

In this paper, genetic algorithm (GA) is applied to the optimum design of reinforced concrete liquid retaining structures, which comprise three discrete design variables, including slab thickness, reinforcement diameter and reinforcement spacing. GA, being a search technique based on the mechanics of natural genetics, couples a Darwinian survival-of-the-fittest principle with a random yet structured information exchange amongst a population of artificial chromosomes. As a first step, a penalty-based strategy is entailed to transform the constrained design problem into an unconstrained problem, which is appropriate for GA application. A numerical example is then used to demonstrate strength and capability of the GA in this domain problem. It is shown that, only after the exploration of a minute portion of the search space, near-optimal solutions are obtained at an extremely converging speed. The method can be extended to application of even more complex optimization problems in other domains.